Regarding graphs - there's nothing much mathematical about them. It's
just circles and arrows like you'd draw on the board. I obviously need
to think about how to make it as simple in my explanations as it is in
reality...
How much speedup do you think you got?
I didn't intend for there to be any bash-isms. The only things I know
of that I use are $(echo hi) and similar - which I didn't think were
bash-only. If they are, then I should get rid of them.
I haven't looked to see that the things I check in /proc would work on
other machines with /proc, but I'm not looking at anything very
complicated - and I suspect something similar exists.
These two scripts need to be combined - since server processes can also
be client processes. This might be best done in some other language -
but I surely hate to create dependencies on having to have a specific
version of perl or python on every machine... Just like I hate having to
have bash on every machine.
In the end this should all run on any POSIX machine. Solaris, *BSD,
AIX, HP-UX, etc.
Although Linux is all I personally care much about, I know other
customers run other OSes - even Windows [which will need its own
separate methods].
Nevertheless, I much appreciate your efforts, and will take a look at
your code - probably this weekend.
This makes you the first provider of a patch to the project!
Awesome!
Thanks!
On 07/31/2012 09:56 AM, David Klann wrote:
> On Mon, 30 Jul 2012 22:57:05 -0600 you coerced some electrons to say:
>
>> I wrote a new blog post on the new code that does client, server, and
>> *dependency* discovery with the dependency data going into our graph
>> database. This is all done via "Stealth" discovery - with no packets
>> going onto the network to discover anything.
>>
>> And it all works!!
>>
>> The blog site ishttp://techthoughts.typepad.com/ with this
>> particular article being here:
>> http://bit.ly/M7jdq0
>>
>> At the end of the post is a graph of all the clients and servers on
>> my machine - it's rather large, so you'll have to click on it to see
>> any detail ;-)
>>
>> If you have questions, don't hesitate to ask. There are no dumb
>> questions -- but those who know me know that I have plenty of dumb
>> answers ;-)
>>
>>
>> As always, looking for feedback.
>>
>> -- Alan Robertson
> Hey Alan,
>
> This looks very cool. I don't have enough of a math background to
> understand network graphs and such, but I've been exposed to the
> concept for enough years that I "kind of" get it.
>
> I'd like to help with a feeble contribution. I've gone through your
> shell scripts and tweaked a bit for speed. Patches for
> discovery_agents/tcplisteners and discovery_agents/tcpclients are
> attached. My sole assumption is that /bin/sh == /bin/bash. There were
> already some bash-isms in there, so this seems a safe assumption. ZSH
> has other capabilities that could get you even better optimization, but
> it's not as widespread as bash (yet! :).
>
> And please ignore all this if you consider the two shell scripts as
> simply a sandbox for tools that will ultimately be written in a
> language that delivers even better performance. This would be a great
> job for a Python or Perl script.
>
> Thanks for all the work you're doing on this!
>
>
> ~d
--
Alan Robertson<[email protected]> - @OSSAlanR
"Openness is the foundation and preservative of friendship... Let me claim
from you at all times your undisguised opinions." - William Wilberforce
_______________________________________________
Assimilation mailing list - Discovery-Driven Monitoring
[email protected]
http://lists.community.tummy.com/cgi-bin/mailman/listinfo/assimilation
http://assimmon.org/
